A bayesian inference of material parameters from DIC data

Mainak Bhattacharyya (1) , Pierre Feissel (1)

1 Roberval - Roberval

The framework of the study is to obtain deterministic and probabilistic approximations of the material parameters from measurement informations acquired from digital image correlation data. The deterministic process involves formulation of an optimal control approach. The probabilistic framework inculcates this optimal control method in a Bayesian inference framework. A Markov Chain Monte Carlo sampling method is applied to obtain the posterior probability density function, along with a radial basis function type interpolation for the numerical frugality of the samplings.

Mots clés: inverse problem; identification; Bayesian inference.

Mots clés: approche inverse; mesure de champ; inférence bayésienne; incertitude

Domaines: Mécanique des structures [physics.class-ph]

Mainak Bhattacharyya, Pierre Feissel. A bayesian inference of material parameters from DIC data. 15ème colloque national en calcul des structures, Université Polytechnique Hauts-de-France [UPHF], May 2022, 83400 Hyères-les-Palmiers, France. ⟨hal-03717570⟩